Abstract The advancement of cancer epigenetics relies on highly efficient methods to detect DNA methylation. Our goal in this project is to devise a novel methodology for high-efficiency and high-throughput DNA methylation profiling by whole genome bisulfite sequencing (WGBS) at the single cell level. Development of such methods to profile single cells will directly impact our understanding of tumor heterogeneity, and the origin, evolution and roles of DNA methylation alterations during cancer initiation and progression. Most importantly, these methods will have direct impact in clinical management of cancer research by enabling genome-wide methylation profiling of limiting tissue biopsies and circulating tumor cells (CTCs). The need for such methodologies is highlighted by recent high-profile publications reporting approaches for single-cell methylation profiling1-3. However, these currently published approaches, which are largely based on ad hoc RNA-seq methodologies, have very low cytidine-guanosine dinucleotide (CpG) coverage (10-20%) and require multiple steps to process the single-cell DNA. Thus the methodologies for single-cell methylation profiling need substantial improvements, as emphasized in a recent review which lists the sparsity of CpG coverage as the major issue impeding use of current approaches4. Our proposed methodology employs innovative modifications: (a) a modified bisulfite (BS) treatment protocol using silica superparamagnetic-beads; (b) novel primer designs for efficient whole genome amplification (WGA) of BS-treated DNA resulting in high-molecular weight double-stranded DNA (dsDNA); (c) implementation of WGA in water-in-oil emulsion generated using microfluidics; (d) Tn5 transposase mediated adapter tagging of dsDNA for sequencing-library preparation. Our proposal leverages our previous work on the bisulfite conversion of low amounts of DNA using magnetic-beads, combined with our preliminary work towards simplified and highly efficient post-BS genome amplification of very low amounts of DNA. These innovative modifications will be streamlined for high-throughput and efficient sample processing to capture majority of the target DNA in the sequencing library, which we expect will substantially improve the coverage of the CpGs across the genome to 300-500% over current methodologies, resulting in an exponential improvement in comparative power. Development of the proposed technique will thus provide, for the first time, direct comparison of epigenetic heterogeneity in as few as two cells. Finally, we will validate and demonstrate the utility of our innovative methodology for mapping the earliest epigenetic changes during tumor initiation using our recently developed inducible model of ex vivo tumorigenesis in colon-derived organoids.